1. Field of the Invention
The present invention generally relates to computer implemented planning resources and decision support tools and, more particularly, to a tool in which production planning information is used to match assets to demands.
2. Background Description
In a complex manufacturing environment such as semiconductor manufacturing, it is very difficult to predict the availability of components/parts. Currently existing planning tools used for predicting availability include spreadsheets and manual calculations. Most planners use a spreadsheet to combine supply and demand information for managing parts and components, and for making reasonably accurate product forecasts.
However, at each production stage, there may be several byproducts or production options that complicate the whole manufacturing scenario. So, in addition to which option to select, product planners must consider which byproduct might be produced. Tracking all projected information receipts with a spreadsheet, while simultaneously attempting to minimize byproduct manufacture is both time consuming and very likely, not very accurate. Further, manually doing the same task is even more difficult. Thus, regardless of which prior art method used, it is nearly impossible to achieve an optimum solution.
Further, most current tools use calculator like logic to predict part availability for each individual time period. Typically, part availability is determined by subtracting Demand and Reservations from Total Supply. However, the total available supply either must be known or, projected supply must be entered, manually. Unfortunately, this data is not available, typically, and so, the above availability check yields a zero result. Thus, due to lack of data, these simple tools cannot give a projected delivery date.
So, to project part availability dates, more advanced, state of the art tools either use part production cycle time for projections or, rely on historic information. Still, in a large and complex manufacturing environment, these state of the art tools fall short, resulting either in a sub-optimal solution or predicting an erroneous availability date.
The tables of FIG. 1 show a simple behavioral example of a typical traditional prior art availability system. The system of the example of FIG. 1 includes a table 50for supply, a table 52for demand and a table 54indicating excess capacity, i.e., the capacity available to promise (ATP). From supply table50, 100 pieces are produced during each production period, e.g., every day. From demand table 52, there is a firm order in Period 2 of 300 pieces. Thus, the supply from Period 1, Period 2 and Period 3  are dedicated to satisfy that firm order of 300 pieces. So, the ATP table 54 shows 100 pieces available in each of Period 4 and Period 5. Consequently, to respond to a new request, the best recommendation from this prior art system is to fill the new request with parts from Period 4 and/or Period 5.
This example of prior art availability allocation represents what is known as Demand Prioritization. The ATP quantity is determined by netting demand against supply. Thus, current demands are filled first from available supply with remaining excess supply being tagged as ATP, i.e., Available To Promise.
In traditional prior art ATP systems, all customers are treated equally and, new commitments are made on a first come first served basis. These prior art systems did not make distinctions between requests from a higher priority customers (e.g., tier 1) and lower priority customers (tier 3). Further, system users of these prior art systems did not have any capability of prioritizing demands, giving higher priority to higher tier customers' demand. Consequently, demand from lower tier customers could not be overridden by demands from higher tier customers.
Additionally, capacity could not be reserved for important customers on these prior art systems, which could not pre-allocate or reserve portions of output for a particular customer. Typically, for these prior art systems, order commitment was based on Supply Determination, i.e., determining supply coming to stock or, supply already available as inventory. The common prior art supply determination models are: Setting supply equal to the forecasted demand; and Basing end product availability on sub-components availability, i.e., those sub-components used in producing the product supplied.
Unfortunately, the disadvantage of the former method of basing supply on demand is that output or supply is based on a demand forecast and, therefore, is only as good as the forecast. The disadvantage of the latter method of basing supply availability on sub-component availability is that it results in myopic Inventory allocation, failing to take into account the business impact of each allocation.
A standard semiconductor industry practice called “Binning” makes this latter prior art method especially unsuitable for semiconductor manufacturing. Where manufacturing a part provides variations of the part due to normal manufacturing parameter variations, the parts may be sorted according into separately identified groups, e.g., fast parts, slow parts and normal parts. This is referred to as binning.
As a result of normally unpredictable manufacturing process variations, the number of parts falling into each “bin” varies from production lot to production lot. So, one lot may produce fewer fast parts and more normal parts than expected, while another lot may produce fewer than expected slow parts. Satisfying demand using binned sub-components, for this latter ATP method, may use more sub-components than necessary. As a result, projected ATP is lower than what the enterprise is actually capable of producing.
In some instances shortfalls may be avoided where, for example faster parts may be substituted, transparently, for slower parts. Further, such substitutions may be time-dependent, i.e., based on business decisions, allowing, over time, changing or halting the substitution. Prior art ATP systems could not handle these on the fly decisions.
Furthermore, the manufacturer may resort to second sourcing or out sourcing for the same part. Also, alternate processes may be modeled to reflect alternate bills of material resulting in different sub-components, which these prior art ATP systems also cannot handle. Instead, these prior art ATP systems ignore manufacturing capacity and consider only sub-component availability.
Thus, basing commit orders on supply availability, prior art ATP systems require manual intervention, because these prior art systems are unable to recognize important customers and allocate supply, however limited, accordingly. Consequently the prior art ATP system user must compensate for prior art ATP system shortcomings.
For example, a customer order may have multiple delivery dates. The prior art ATP system user must decide whether to complete the order with supply available on each requested date; determine the latest date when the entire supply will be available and commit to that date; or, fill the order according to the schedule proposed by ATP system. The prior art ATP system user must make this decision for each individual order.
Another limitation of prior art ATP systems is that they are connected, generally, to a single ordering system, and are not capable of accepting orders from multiple lines. This lack of horizontal integration results in potential commit errors because all demands from other ordering system are not reflected in a single common ORDER BOOK. Consequently, there may be system downtime to periodically synchronize supply and demand as required in these prior art ATP systems. This synchronization is done, for example, using batch programs.
Another shortfall of these prior art ATP systems is that supply generation is not linked, explicitly, with demands and requests on hand. Consequently, very often this results in excess part supply with no demand, and with limited or no availability for high demand parts.
Thus, there is a need for an integrated ATP system capable receiving orders from multiple ordering systems, providing a scheduled date of when ordered material can be supplied and the quantity that can be supplied on that date in response to the customer's preselected requirements.